OpenAI and Anthropic representatives pledge faster safety incident disclosures at Australian hearing · Digg